| Space-time adaptive processing (STAP) techniques developed over the last thirtyyears have made great achievements in theory, but some problems still need to besolved. Generally, when the airborne radar looks downwards, the clutter environmentsare nonhomogeneous. In the nonhomogeneous environments, the statistical properties ofthe secondary data are different, and the independent and identically distributedsecondary data cannot be obtained sufficiently, which make the covariance matrixestimation error increase and the performance of the system fall. Some new solvablemethods such as forwards-backwards averaging, direct data domain method andparametric method are presented, which can improve the performance of the system to acertain extent. However, in order to improve the performance further, researchers attachimportance to KA-STAP (Knowledge-aided STAP) technique gradually. Recently, withthe development of all kinds of digital terrain data and artificial constructions’ identifytechniques, KA-STAP techniques using digital terrain data have obtained more andmore study.In this dissertation, the properties of two-dimensional space-time clutter spectrumand the principle of STAP are introduced. Through the analysis, fully adaptive STAP isan optimum processing configuration, but the computation load and the need foradequate homogeneous training data limit its application. Reduced-dimension STAP canovercome these problems and it is applied in engineering extensively. Then weintroduce and analyze several common reduced-dimension STAP methods. At last, theKA-STAP method using terrain classification data and terrain elevation data isintroduced in detail, and simulation shows that KA-STAP using digital terrain data canimprove the performance of the system. Because the radar data processing and digitalterrain data generally adopt different coordinates, the concepts of several coordinatesand the coordinates converting methods are also introduced detailedly. |
PDF Full Text Request